A wind turbine generates power by rotating the main shaft connected to blades subjected to winds, increasing the rotation of the main shaft with the gearbox, and rotating the rotor of the power generator. Technologies as described below are known as abnormality diagnosing apparatuses that diagnose an abnormality of the main shaft hearing, the gearbox, the power generator, or the like.
Japanese Patent Laying-Open No. 2009-243428 (PTD 1) discloses an apparatus for monitoring a wind turbine. This apparatus for monitoring a wind turbine monitors the condition of the wind turbine by using characteristic values generated based on measurement data measured by a plurality of sensors provided in the wind turbine. The monitoring apparatus includes first storage means that stores a plurality of characteristic values associated with a measured time for each characteristic item, wherein, when characteristic values associated with the same measured time are defined as a single data set, the data set is stored with identification information indicating a classification determined in accordance with a characteristic value of a predetermined characteristic item; second storage means that stores a plurality of characteristic values associated with a measured time for each characteristic item, wherein, when characteristic values associated with the same measured time are defined as a single data set, the data set is stored with identification information indicating a classification determined in accordance with a characteristic value of a predetermined characteristic item, and the characteristic value of the specific characteristic item forming the data set falls within a predetermined reference range defined in advance; diagnostic setting means that extracts and sets a plurality of data sets to be used for a diagnosis from the first storage means, and extracts and sets a plurality of data sets to be used for the diagnosis from the second storage means; index value calculating means that calculates a condition index value that represents the condition of the wind turbine, by using a statistical calculation technique, based on the data sets of a data file to be diagnosed set by the diagnostic setting means and data sets of a reference data file; evaluating means that evaluates the condition of the wind turbine based on the condition index value calculated by the index value calculating means; and notifying means that notifies a result evaluated by the evaluating means.
This monitoring apparatus can automatically monitor the condition of the wind turbine, and quantitatively evaluate the condition based on an appropriate reference (see Japanese Patent Laying-Open No. 2009-243428).